Transcript Nonlinear DOAS - uni

Contribution of stratospheric cooling to
satellite-inferred tropospheric temperature
trends
Quiang Fu, Celeste M. Johanson, Stephen G. Warren & Dian J. Seidel
NATURE | VOL 429 | 6 may 2004
pp. 55-58
Outline
• The Microwave sounding unit (MSU)
Measurement principle
• Tropospheric temperature trends derived from
satellites
• Discrepancies between surface and satellite
measurements
• Reason for popular and scientific discussions about
climate change
• Reconciliation by this paper…
MSU
• nadir viewing instrument
• 4 microwave channels (50-57.9 GHz)
• measures brightness temperature according to
Planck‘s law:
Radiative transfer (no scattering)
• Simplified radiative transfer (Schwarzschild
equation)
Weighting function
Contribution from the atmosphere
Measured intensity
Surface Contribution
Spectral range
• MSU measures at 50.3, 53.74, 54.96, and
57.95 GHz
• Oxygen absorption in this region
• Each channel has a different optical depth
with respect to oxygen
RT II
Channel 2 -> troposphere Optically thin
Optically thick
Results from MSU
Troposphere
Stratosphere
(UAH = University at Huntsville, RSS = Remote Sounding Systems)
Results II
Results II
Discrepancies
• Surface measurements:
0.17K/decade
• MSU (UAH):
0.047K/decade
• MSU (RSS):
0.13K/decade
• GCM prediction for troposphere:
1.1*surface temperature change (0.187
K/decade) globally
 Boost for climate change critics…
The solution by Fu et al
There is still an impact of the
stratosphere to T2 (15%)
and the stratosphere is cooling!!
Correction for T2 necessary:
T850-300=a0+a2T2+a4T4
a‘s derived from fits to ballon borne
measurements
Dependencies of T2 and T4
T4
Stratospheric trend
T2
Tropospheric trend
0.18K
Tropospheric trend
Results
Results
Trend ratios
(troposphere/surface)
Globe
GCM MSU
1.2
1.1
Tropics 1.54
1.6
Conclusion
• Wrong troposheric temperature trend due to the
disregard of stratospheric cooling
• Simple correction scheme was implemented
• New results reconcile surface measurements, MSU
measurements and GCM models
• ASMU provides now more channels (better height
resolution)
• Always look on the averaging kernel (esp. for
NADIR DOAS measurements) !!